U.S. patent application number 11/289184 was filed with the patent office on 2008-02-07 for efficient cell selection.
Invention is credited to Magnus Almgren, Henrik Asplund, Anders Furuskar, Bengt Lindoff, Niclas Wiberg.
Application Number | 20080031368 11/289184 |
Document ID | / |
Family ID | 37943933 |
Filed Date | 2008-02-07 |
United States Patent
Application |
20080031368 |
Kind Code |
A1 |
Lindoff; Bengt ; et
al. |
February 7, 2008 |
Efficient cell selection
Abstract
Systems and methods of cell selection are based on measures of
frequency selectivity, time selectivity, and/or antenna
selectivity. For example, the delay spread and coherence bandwidth
of a channel are two measures of frequency selectivity that can be
taken into account in the cell selection process. In addition to or
instead of either delay spread or coherence bandwidth, Doppler
frequency shift or coherence time (which are two measures of time
selectivity) and/or antenna correlation or mean signal strength per
antenna (which are two measures of antenna selectivity), among
other parameters, can be determined and taken into account.
Inventors: |
Lindoff; Bengt; (Bjarred,
SE) ; Almgren; Magnus; (Sollentuna, SE) ;
Asplund; Henrik; (Stockholm, SE) ; Furuskar;
Anders; (Stockholm, SE) ; Wiberg; Niclas;
(Linkoping, SE) |
Correspondence
Address: |
POTOMAC PATENT GROUP PLLC
P. O. BOX 270
FREDERICKSBURG
VA
22404
US
|
Family ID: |
37943933 |
Appl. No.: |
11/289184 |
Filed: |
November 29, 2005 |
Current U.S.
Class: |
375/260 ;
455/450 |
Current CPC
Class: |
H04W 48/20 20130101;
H04B 17/382 20150115; H04B 17/318 20150115; H04L 27/2601 20130101;
H04B 17/336 20150115 |
Class at
Publication: |
375/260 ;
455/450 |
International
Class: |
H04J 11/00 20060101
H04J011/00; H04Q 7/20 20060101 H04Q007/20 |
Claims
1. An apparatus in a communication system that includes a plurality
of cells, wherein the cells transmit respective signals that
include respective pilots through respective channels to a
terminal, the apparatus comprising: a signal strength estimator
configured to generate first estimates of at least one of signal
strengths and signal-to-interference ratios of signals received
from respective cells; an estimator configured to generate second
estimates of at least one of a frequency selectivity, a time
selectivity, and an antenna selectivity of respective channels
between the terminal and respective cells; and a cell selector
configured to generate a cell selection based on the first
estimates and the second estimates.
2. The apparatus of claim 1, wherein second estimates of the
frequency selectivity of a channel include estimates of at least
one of a coherence bandwidth and a delay spread of the channel.
3. The apparatus of claim 2, wherein the signals are orthogonal
frequency division multiplex signals and the estimator is
configured to generate second estimates by calculating a
correlation in signal strength between pilots at different
frequencies and at a same time instant.
4. The apparatus of claim 1, wherein second estimates of the time
selectivity of a channel include estimates of at least one of a
Doppler frequency shift and a coherence time of the channel.
5. The apparatus of claim 4, wherein the signals are orthogonal
frequency division multiplex signals and the estimator is
configured to determine correlations in signal strength between
pilots at different time instants and at a same frequency for
respective cells.
6. The apparatus of claim 1, wherein second estimates of the
antenna selectivity of a channel include estimates of an antenna
correlation of the channel.
7. The apparatus of claim 1, wherein the signals are orthogonal
frequency division multiplex signals and the estimator is
configured to determine correlations in signal strength between
pilots from different antennas at a cell and at same time instants
and same frequencies.
8. The apparatus of claim 1, wherein the signal strength estimator
is configured to generate mean signal levels of pilots received
from cells.
9. The apparatus of claim 8, wherein the signal strength estimator
is configured to filter the mean signal levels generated.
10. The apparatus of claim 1, wherein the cell selector generates
the cell selection by computing a cell selection function:
Cell=f(S.sup.i, T.sub.d.sup.i, F.sub.D.sup.i, C.sub.a.sup.i) for
each cell, in which Cell represents a selected cell, S.sup.i
represents a first estimate for a cell i, T.sub.d.sup.i represents
a delay spread of the channel between the cell i and the terminal,
F.sub.D.sup.i represents a Doppler frequency shift of the channel
between the cell i and the terminal, and C.sub.a.sup.i represents
an antenna correlation of the channel between the cell i and the
terminal.
11. The apparatus of claim 10, wherein the cell selector is
configured to determine a quality metric for each cell, and the
cell selection function determines the cell i having the largest
quality metric.
12. The apparatus of claim 1, wherein the cell selector generates
the cell selection based on the first estimates of the
signal-to-interference ratio and the second estimates.
13. The apparatus of claim 1, wherein the apparatus is included in
the terminal.
14. The apparatus of claim 1, wherein the first and second
estimates are generated in the terminal, and the cell selector is
included in at least one other entity in the communication
system.
15. A method of cell selection in a communication system that
includes a plurality of cells, wherein the cells transmit
respective signals that include respective pilots through
respective channels to a terminal, the method comprising:
generating first estimates of at least one of signal strengths and
signal-to-interference ratios of signals received from respective
cells; generating second estimates of at least one of a frequency
selectivity, a time selectivity, and an antenna selectivity of
respective channels between the terminal and respective cells; and
selecting a cell based on the first estimates and the second
estimates.
16. The method of claim 15, wherein generating second estimates of
the frequency selectivity of a channel includes generating
estimates of at least one of a coherence bandwidth and a delay
spread of the channel.
17. The method of claim 16, wherein the signals are orthogonal
frequency division multiplex signals and the second estimates are
generated by calculating a correlation in signal strength between
pilots at different frequencies and at a same time instant.
18. The method of claim 15, wherein generating second estimates of
the time selectivity of a channel includes generating estimates of
at least one of a Doppler frequency shift and a coherence time of
the channel.
19. The method of claim 18, wherein the signals are orthogonal
frequency division multiplex signals and the second estimates are
generated by determining correlations in signal strength between
pilots at different time instants and at a same frequency for
respective cells.
20. The method of claim 15, wherein generating second estimates of
the antenna selectivity of a channel includes generating estimates
of an antenna correlation of the channel.
21. The method of claim 15, wherein the signals are orthogonal
frequency division multiplex signals and the second estimates are
generated by determining correlations in signal strength between
pilots from different antennas at a cell and at same time instants
and same frequencies.
22. The method of claim 15, wherein generating first estimates
includes generating mean signal levels of pilots received from
cells.
23. The method of claim 22, wherein generating first estimates
includes filtering the mean signal levels generated.
24. The method of claim 15, wherein selecting the cell includes
computing a cell selection function: Cell=f(S.sup.i, T.sub.d.sup.i,
F.sub.D.sup.i, C.sub.a.sup.i) for each cell, in which Cell
represents a selected cell, S.sup.i represents a first estimate for
a cell i, T.sub.d.sup.i represents a delay spread of the channel
between the cell i and the terminal, F.sub.D.sup.i represents a
Doppler frequency shift of the channel between the cell i and the
terminal, and C.sub.a.sup.i represents an antenna correlation of
the channel between the cell i and the terminal.
25. The method of claim 24, wherein selecting the cell includes
determining a quality metric for each cell, and the cell selection
function determines the cell i having the largest quality
metric.
26. The method of claim 15, wherein selecting the cell includes
computing the cell selection based on signal-to-interference ratios
and the second estimates.
27. The method of claim 15, wherein the method is carried out in
the terminal.
28. The method of claim 15, wherein the steps of generating the
first and second estimates are carried out in the terminal; further
comprising the step of communicating the first and second estimates
to at least one other entity in the communication system; the step
of selecting the cell is carried out in the at least one other
entity; and further comprising the step of communicating the
selected cell to the terminal.
Description
BACKGROUND
[0001] This invention relates to communication systems and more
particularly to digital communication systems.
[0002] Third generation (3 G) cellular wireless communication
systems based on wideband code division multiple access (WCDMA)
technology are being deployed all over the world. These systems are
standardized by specifications promulgated by the Third Generation
Partnership Project (3 GPP). Evolution of WCDMA radio access
technology has occurred with the introduction of high-speed
downlink packet access (HSDPA) and an enhanced uplink (UL).
[0003] FIG. 1 depicts a typical cellular wireless telecommunication
system 10. Radio network controllers (RNCs) 12, 14 control various
radio network functions, including for example radio access bearer
setup, diversity handover, etc. In general, each RNC directs calls
to and from a mobile station (MS), or remote terminal or user
equipment (UE), via the appropriate base station(s) (BSs), which
communicate with each other through downlink (DL) (i.e.,
base-to-mobile or forward) and UL (i.e., mobile-to-base or reverse)
channels. In FIG. 1, RNC 12 is shown coupled to BSs 16, 18, 20, and
RNC 14 is shown coupled to BSs 22, 24, 26.
[0004] Each BS, or Node B in 3 G vocabulary, serves a geographical
area that is divided into one or more cell(s). In FIG. 1, BS 26 is
shown as having five antenna sectors S1-S5, which can be said to
make up the cell of the BS 26, although a sector or other area
served by signals from a BS can also be called a cell. In addition,
a BS may use more than one antenna to transmit signals to a UE. The
BSs are typically coupled to their corresponding RNCs by dedicated
telephone lines, optical fiber links, microwave links, etc. The
RNCs 12, 14 are connected with external networks such as the public
switched telephone network (PSTN), the internet, etc. through one
or more core network nodes, such as a mobile switching center (not
shown) and/or a packet radio service node (not shown).
[0005] In a communication system such as that depicted in FIG. 1,
each BS usually transmits predetermined pilot symbols on the DL
physical channel (DPCH) to a UE and on a common pilot channel
(CPICH). A UE typically uses the CPICH pilot symbols in deciding
which BS to listen to, which is a process called cell selection,
and in estimating the impulse response of the radio channel to the
BS. It will be recognized that the UE uses the CPICH pilots for
channel estimation, rather than the DPCH pilots, due to the CPICH's
typically higher signal-to-noise ratio (SNR). The UE uses the DPCH
pilots for estimation of the signal-to-interference ratio (SIR),
i.e., for DL transmission power control, among other things.
[0006] As UEs move with respect to the BSs, and possibly vice
versa, on-going connections are maintained through a process of
handover, or hand-off, in which as a user moves from one cell to
another, the user's connection is handed over from one BS to
another. Early cellular systems used hard handovers (HHOs), in
which a first cell's BS (covering the cell that the user was
leaving) would stop communicating with the user just as a second BS
(covering the cell that the user was entering) started
communication. Modern cellular systems typically use diversity, or
soft, handovers (SHOs), in which a user is connected simultaneously
to two or more BSs. In FIG. 1, MSs 28, 30 are shown communicating
with plural BSs in diversity handover situations. MS 28
communicates with BSs 16, 18, 20, and MS 30 communicates with BSs
20, 22. A control communication link between the RNCs 12, 14
permits diversity communications to/from the MS 30 via the BSs 20,
22.
[0007] New radio transmission technologies are being considered for
evolved-3 G and fourth generation (4 G) communication systems,
although the structure of and functions carried out in such systems
will generally be similar to those of the system depicted in FIG.
1. In particular, orthogonal frequency division multiplexing (OFDM)
is under consideration for evolved-3 G and 4 G systems. An OFDM
system can adapt its DL transmission parameters not only in the
time domain, as in current communication systems, but also in the
frequency domain. This can provide higher performance where the DL
communication channel varies significantly across the system
bandwidth. For example, combined time- and frequency-domain
adaptation may yield a capacity gain of a factor two compared to
time-domain-only adaptation for a so-called 3 GPP Typical-Urban
channel and a system bandwidth of 20 megahertz (MHz).
[0008] As described above, cell selection and handover are
fundamental functions in cellular communication systems in that
these functions determine which cell(s) a remote terminal
communicates with. The terms "cell selection" and "handover" are
sometimes given distinguishable meanings. For example, "cell
selection" can refer to a function in an idle terminal and
"handover" can refer to a function in an active terminal.
Nevertheless, the term "cell selection" is used in this application
to cover both functions for simplicity of explanation.
[0009] Cell selection has a number of objectives, which include
connecting terminals to the cell(s) that will provide the highest
quality of service (QoS), consume the least power, and/or generate
the least interference. It is also of interest to make robust cell
selections, thereby limiting the number and frequency of cell
re-selections.
[0010] Cell selection is traditionally based on the signal strength
or SNR of candidate cells. For example, U.S. patent application
Ser. No. ______ filed on Nov. 29, 2005, by B. Lindoff for "Cell
Selection in High-Speed Downlink Packet Access Communication
Systems", which is incorporated here by reference, describes a cell
selection process that also takes into account the delay spread of
the communication channel. For a given SNR, different delay spreads
yield different qualities of service (e.g., different bit rates),
and by taking this into account in the cell selection procedure,
improved QoS can be achieved. In that patent application, the path
delay profile in a typical WCDMA communication system is described
as a useful representation of the delay spread.
[0011] It seems unlikely that estimation of the delay spread in an
OFDM communication system would be done in the same way as in a
WCDMA system. Moreover, the delay spread does not capture all of
the variability of the communication channel, which also arises
from the mobility of the UE and relay nodes or BSs with respect to
one another, and from the correlation properties of signals
transmitted from different antennas. Highly correlated antennas,
which is to say antennas that produce signals that are highly
correlated, yield little diversity gain, and so such antennas
result in greater signal variations at receivers, leading to
decreased cell selection accuracy. Correlation functions and their
use in characterizing communication channels such as those in
cellular communication systems are described in J. Proakis,
"Digital Communications", Section 14.1.1, 4th ed., McGraw-Hill
(2001).
SUMMARY
[0012] This application describes systems and methods of cell
selection that do not suffer from these and other problems with
prior systems and methods. In contrast with prior systems and
methods, cell selection is based on measures of frequency
selectivity, time selectivity, and/or antenna selectivity. For
example, the delay spread and coherence bandwidth of a channel are
two measures of frequency selectivity that can be taken into
account in the cell selection process. In addition to or instead of
either delay spread or coherence bandwidth, Doppler frequency shift
or coherence time (which are two measures of time selectivity)
and/or antenna correlation or mean signal strength per antenna
(which are two measures of antenna selectivity), among other
parameters, can be determined and taken into account.
[0013] In one aspect of this invention, there is provided an
apparatus in a communication system that includes a plurality of
cells, which transmit respective signals that include respective
pilots through respective channels to a terminal. The apparatus
includes a signal strength estimator configured to generate first
estimates of at least one of signal strengths and
signal-to-interference ratios of signals received from respective
cells; an estimator configured to generate second estimates of at
least one of a frequency selectivity, a time selectivity, and an
antenna selectivity of respective channels between the terminal and
respective cells; and a cell selector configured to generate a cell
selection based on the first estimates and the second
estimates.
[0014] In another aspect of this invention, there is provided a
method of cell selection in a communication system that includes a
plurality of cells, which transmit respective signals that include
respective pilots through respective channels to a terminal. The
method includes generating first estimates of at least one of
signal strengths and signal-to-interference ratios of signals
received from respective cells; generating second estimates of at
least one of a frequency selectivity, a time selectivity, and an
antenna selectivity of respective channels between the terminal and
respective cells; and selecting a cell based on the first estimates
and the second estimates.
BRIEF DESCRIPTION OF THE DRAWINGS
[0015] The several features, advantages, and objects of this
invention will be understood by reading this description in
conjunction with the drawings, in which:
[0016] FIG. 1 depicts a cellular radio communication system;
[0017] FIG. 2 depicts a time and frequency arrangement of
sub-carriers in a communication system using orthogonal frequency
division multiplexing;
[0018] FIG. 3 depicts a time and frequency arrangement of
sub-carriers that include pilot signals;
[0019] FIG. 4A is a block diagram of a portion of a user equipment
for a communication system;
[0020] FIG. 4B is a block diagram of a portion of a user equipment
and a portion of a communication system; and
[0021] FIG. 5 is a flow chart of a method of cell selection.
DETAILED DESCRIPTION
[0022] The DL in evolved-3 G and 4 G communication systems may be
based on OFDM, by which data is transmitted over a relatively large
set of relatively narrow sub-carriers that are allocated at
different frequencies. This application focusses on OFDM systems
for economy of explanation, but it will be understood that the
principles described in this application can be implemented in
other digital communication systems.
[0023] A basic time-frequency structure of a DL in an OFDM system
is depicted in FIG. 2, which shows a plurality of OFDM sub-carriers
that are contiguous in the frequency direction. The radio resource
devoted to a particular user may be called a "chunk", which is a
particular number of particular sub-carriers used for a particular
period of time. Different groups of sub-carriers are used at
different times for different users, and FIG. 2 illustrates chunks
for four users A, B, C, D. In the downlink of the exemplary OFDM
system depicted by FIG. 2, a chunk includes 15 sub-carriers (not
all of which are shown, for clarity) spaced apart by 13.75
kilohertz (kHz), which together occupy approximately 200 kHz in
frequency, and 0.625 millisecond (ms) in time. It will be
understood that the arrangement of FIG. 2 is just an example and
that other arrangements can be used.
[0024] For cell selection purposes, reference signals, so-called
pilots, can be transmitted from each base station at known
frequency and time instants. An exemplary time-frequency structure
with eight such pilots 302 is depicted in FIG. 3, which shows eight
sub-carriers having the pilots 302 in the OFDM time-frequency
plane. Other OFDM sub-carriers 304 transport data, but for clarity
these are indicated in FIG. 3 at only one instant in the
time-frequency plane. It will be understood that each chunk
typically includes a few pilots on different sub-carriers. It will
also be understood that a BS may use multiple transmit antennas to
transmit information into a cell/sector/area, and those different
transmit antennas may send respective, different pilots.
[0025] In a communication system such as that depicted in FIGS.
1-3, the problems with cell selection that are described above can
be overcome by basing cell selection on at least one of the
frequency selectivity, time selectivity, and antenna selectivity of
the DL channel. Doing so counters effects of multipath propagation,
which results in a pattern of radio waves where minima are
encountered when the vector sum of all waves cancel or almost
cancel. A receiver moving through such a wave pattern experiences
rapid signal variations, or fading, in both the time and frequency
domains that present a challenge to upholding and optimizing the
transmission and reception of information.
[0026] Movement of the receiving antenna through the wave pattern
will result in signal variations in time. By reciprocity, the same
variations will be observed if the direction of transmission is
reversed, i.e., if the moving receiver becomes a moving transmitter
and the stationary transmitter becomes a stationary receiver.
Furthermore, even if both transmitter and receiver are stationary,
movement and changes in the atmosphere and surroundings of the two
may result in changes to the wave pattern and hence time variations
of the received signal. All types of movement give rise to what is
called time-selective multipath fading, or time selectivity, in
this application.
[0027] The phase of each radio wave depends on the path length,
which may be expressed in wavelengths. If the frequency is shifted,
the phase of each radio wave may also be shifted, and the wave
pattern is changed. Thus, at a given time instant, the received
signal will have fading variations over the frequency band, and
this is called frequency-selective multipath fading, or frequency
selectivity, in this application.
[0028] In addition, multiple antennas may be used for transmission
and/or reception of the radio waves. Properties of the antenna
arrangement, such as relative positions, radiation patterns, mutual
coupling, and polarization, result in different weighting and phase
shifts of the radio waves at different antennas. Hence, the wave
pattern associated with one transmitting antenna may be partially
or fully independent of that associated with another transmitting
antenna. By reciprocity, the same holds for different receiving
antennas. Thus, different signal strengths may be encountered by
different antennas, which is called antenna selectivity in this
application.
[0029] For example, Doppler frequency spread and coherence time of
a channel are two measures of time selectivity, the delay spread
and coherence bandwidth of the channel are two measures of
frequency selectivity, and antenna correlation and mean signal
strength per antenna are two measures of antenna selectivity. It
will be appreciated that the time, frequency, and antenna
selectivities can be measured by other parameters. In particular,
it can be advantageous to base an estimate of the delay spread on
the coherence bandwidth of a channel rather than on a path delay
profile. It is known that the delay spread is inversely
proportional to the coherence bandwidth. In some communication
systems, such as OFDM systems, the coherence bandwidth is more
easily measured than delay spread. For example, in OFDM-based
systems, the coherence bandwidth can be easily obtained by
correlating sub-carrier signal strengths.
[0030] FIG. 4A is a block diagram of a UE, such as a mobile
terminal for an OFDM communication system, that includes an
apparatus 400 in accordance with this invention. For simplicity,
only some parts of the UE are shown in the figure. In particular,
signals transmitted by base stations or other entities in the
communication system are received by an antenna 402, which may
include multiple antenna elements, and are down-converted to
base-band signals by a suitable front-end receiver (Fe RX) 404.
Signals from the Fe RX 404 are provided to a suitable detector 406
that produces, e.g., by decoding, information carried by the
signals that may then be further processed by the UE.
[0031] The apparatus 400 includes a signal strength (SS) estimator
408, which generates, based on signals from the Fe RX 404 and on a
regular basis for each detected cell i, an estimate of the cell's
respective signal strength S.sup.i. A suitable SS estimate S.sup.i
is the mean signal level of the pilots received from a BS over the
whole frequency band. Although one embodiment may use the mean of
the signal level over the entire frequency band as a SS measure,
other ways of estimating the signal strength are known, and any of
these can be used. For example, either the minimum or the maximum
pilot signal strength over the band can be used as the SS estimate
S.sup.i. It will be appreciated, however, that it is not necessary
to do any averaging at all, although averaging gives less
variability in the signal strength estimate, which of course is
desirable. Averaging measurements on even a single sub-carrier can
be enough with efficient filtering over fading variations in time.
At the other extreme, the signal strength can be estimated by
averaging over both time and frequency (and antennas, if there are
more than one).
[0032] The estimator 408 optionally can filter the signal level
measurements with either predetermined filter parameters (e.g.,
time constant, etc.) or filter parameters that depend on network
parameters. For example, the filtering can be a moving average
formed with a sliding window of between about 100 ms and a few
hundred milliseconds in width. In systems in which fast cell
selections are desirable, for example systems providing sector
selection, shorter time windows, on the order of milliseconds,
could be used. For another example, the filter may be an
exponential filter, such that
S.sup.i(t)=aS.sup.i(t-1)+(1-a)P.sup.i(t), where P.sup.i(t) is the
level of a pilot at time instant t, and a is a network-dependent
filter parameter, e.g., 0.5, 0.25, or 0.125. It will be appreciated
that the estimator 408 can be implemented by a suitably programmed
processor or suitably configured logic circuits.
[0033] Because cell selection can be based on one or more of the
frequency selectivity, time selectivity, and antenna selectivity of
the DL channel, the apparatus 400 may also include an estimator 410
configured to generate estimates of a measure of the frequency
selectivity. As described above, one suitable measure is the
coherence bandwidth B.sub.c.sup.i of the communication channel
between the UE and a respective BS. The device 410 can generate
such estimates by determining the correlation in signal strength
between pilots at different frequencies but at the same time
instant. Also as described above, another suitable measure of the
frequency selectivity is the delay spread T.sub.d.sup.i which the
estimator 410 can determine from the delay spread's functional
relationship to the coherence bandwidth, i.e.,
T.sub.d.sup.i=f(B.sub.c.sup.i). For example, the functional
relationship can be the inverse, where
T.sub.d.sup.i=1/B.sub.c.sup.i.
[0034] The delay spread T.sub.d can also be determined from the
path delay profile (PDP). Methods of determining PDPs are well
known in the art. For example, the PDP can be estimated by
correlating the received signal with a scrambling code for the
transmitting cell and a pilot's channelization code for different
time lags, where the longest time lag has a length corresponding to
a worst-case assumption of the delay spread, e.g., 100 or so chips
of the scrambling code. Then, peaks in the PDP can be determined as
those peaks in the correlation result that have powers greater than
a threshold, e.g., 5% of the highest peak's power. The rest of the
correlation result can then be assumed to indicate no signal.
[0035] The PDP and the frequency correlation function are a Fourier
transform pair, and so the PDP can be simply estimated by, for
example, taking an inverse fast Fourier transform (IFFT) of an
estimate of the frequency correlation function. The delay spread
T.sub.d characterizes the width of the PDP (e.g., the T.sub.d can
be the total width or the "standard deviation", depending on the
definition of delay spread used), and the coherence bandwidth
B.sub.c characterizes the width of the frequency correlation
function (and also can depend on the definition used).
[0036] As described above, the delay spread T.sub.d and coherence
bandwidth B.sub.c have a functional relationship, but the function
depends on the shape of the PDP (time-averaged to smooth fast
fading) or frequency correlation function. Even so, giving a
certain coherence bandwidth imposes a lower bound on the delay
spread (and vice versa), according to the following expression:
T.sub.d.sup.i.gtoreq.C/B.sub.c.sup.i
where C is a constant. For some PDPs, such as an exponentially
decaying profile, the preceding expression is an equality, but this
can not be assumed for most channel realizations that occur in real
systems. Thus, it can be better (e.g., more accurate) to estimate
the delay spread from the PDP rather than directly from the
functional relationship with the coherence bandwidth. Nevertheless,
the delay spread and coherence bandwidth are in general two equally
good measures of frequency selectivity.
[0037] In view of the functional relationship between the coherence
bandwidth and the delay spread, it can be understood that cell
selection can be based on either the coherence bandwidth B.sub.c or
the delay spread T.sub.d as described in more detail below. Of
course, the (time-averaged) PDP or the frequency transfer function
are more descriptive measures but are more difficult to work with
than those two simple numeric measures. Other commonly used
measures of multipath fading variations are the magnitude variation
(max-min), fading depth, fading width, level crossing rate (LCR),
and average duration of fades (ADF). Any of these measures can be
applied both to the frequency-selective fading and to the
time-selective fading. In any event, it will be appreciated that
the frequency sensitivity estimator 410 can be implemented by a
suitably programmed processor or suitably configured logic
circuits.
[0038] The apparatus 400 may also or instead include an estimator
412 configured to generate estimates of a measure of the time
selectivity of the DL channel. As described above, one suitable
measure is the Doppler frequency shift F.sub.D.sup.i of signals
from the respective cell i. The Doppler spread reflects the
relative speed of a terminal and base station or relay node, and a
large Doppler spread generally indicates large channel variations.
Another suitable measure is the coherence time, which can be
estimated by computing the correlation in signal strength of pilots
at different time instants but at the same frequency. The Doppler
frequency shift F.sub.D.sup.i is just the inverse of the coherence
time, and so a way to estimate one is also a way to estimate the
other. The Doppler frequency spread and the coherence time have the
same type of inequality functional relationship described above
with respect to the delay spread and the coherence bandwidth. The
space-time correlation function (auto-correlation function) and the
Doppler spectrum are measures that are more descriptive of time
selectivity than the simpler numeric measures Doppler spread and
coherence time, but can be more difficult for a processor to use.
It will be appreciated that the estimator 412 can be implemented by
a suitably programmed processor or suitably configured logic
circuits.
[0039] The apparatus 400 may also or instead include an estimator
414 configured to generate estimates of a measure of the antenna
selectivity of the DL channel. As described above, one suitable
measure is the antenna correlation C.sub.a.sup.i. Because the UE
can identify which received signals come from which of possibly
several antennas or antenna lobes at a BS or relay node, such an
estimate can be generated by calculating the correlation in signal
strength between pilots from different antennas but at the same
time instants and the same frequencies. Such an estimate can also
be generated by determining the mean signal strengths of the
antennas in each cell i, and it will be understood that the mean
signal strength per antenna and the antenna correlations are two
different measures of the antenna selectivity. The antenna
correlation C.sub.a.sup.i indicates the amount of antenna
diversity, which in turn indicates how much the channel may be
expected to vary. A high antenna correlation C.sub.a.sup.i
indicates little diversity and thus typically large channel
variations. It will be appreciated that the estimator 414 can be
implemented by a suitably programmed processor or suitably
configured logic circuits.
[0040] It will be appreciated that the number of antennas itself
may be a useful measure of antenna selectivity. It is currently
believed that most other measures are usually specific to a
particular antenna arrangement, such as a uniform linear array. The
combination of mean signal strength per antenna and correlations
between all pairs of antennas provides a full description of the
antenna selectivity.
[0041] Information from the estimator 408 and one or more of the
estimators 410, 412, 414 is provided to a cell selector 416, which
generates a cell selection signal based on the signal strength and
at least one of the frequency, time, and antenna selectivities. The
selector 416 may trigger a change of cell either by itself or by
reporting its values to the network. A suitable signal indicating a
change of cell or a value computed by the selector 416 can be
provided to a modulator 418, which also receives other data to be
transmitted. For example, in a communication system such as a WCDMA
system according to the 3 GPP specification, the UE can trigger an
event 1D (change of best cell) by transmitting a Layer-3 radio
resource control (RRC) message. The change signal may be
transmitted to the base station either on occurrence of the event
or on a regular basis. The change signal and data are appropriately
transformed into a modulation signal, which is provided to a
front-end transmitter (Fe TX) 420 that up-converts or otherwise
transforms the modulation signal for transmission to the base
station(s) and other entities in the communication system.
[0042] In general, the cell selector 416 computes a cell selection
function that determines the selected cell, and such a cell
selection function f may have the following general form:
Cell=f(S.sup.i, T.sub.d.sup.i, F.sub.D.sup.i, C.sub.a.sup.i)
[0043] For example, the cell selection function f may be a product,
with the signal strength S.sup.i of each cell multiplied by one or
more respective weight factors. One weight factor applied to the
signal strength S.sup.i can advantageously be an increasing
function of the coherence bandwidth B.sub.c.sup.i. If desired,
other weight factors applied to the signal strength can be a
decreasing function of the Doppler spread F.sub.D.sup.i, and/or a
decreasing function of the antenna correlation C.sub.a.sup.i. It
may be noted that the signal strength S.sup.i and correlation
C.sub.a.sup.i are two measures that complement each other, i.e.,
each does not in itself completely describe the antenna
selectivity.
[0044] It will be appreciated that other exemplary cell selection
functions f and weight-generating functions may be used. For
example, the cell selection function f can be a summation over j=1,
2, . . . , J utility functions, which is to say that:
Cell=Arg.sub.imax{.SIGMA.(f.sub.j(v(i, j), j), i},
which selects that cell i having the largest argument.
[0045] An example of a suitable utility function u(x) is a
piece-wise linear ramping function given by:
u(x)=0, for x<0
u(x)=x, for 0.ltoreq.x.ltoreq.1, and u(x)=1, for x>1.
It will be noted that other utility functions, including linear,
sigmoid/step, and non-linear functions, may be used.
[0046] With such a utility function u, a cell quality metric
Q.sup.i can be computed for each cell i according to:
Q.sup.i=SS.sup.i+K.sub.Tdu(T.sub.d.sup.i/T.sub.dref)+K.sub.Tcu(1-T.sub.c-
.sup.i/T.sub.cref)+K.sub.Cau(C.sub.a.sup.i/C.sub.aref)
where SS.sup.i is the mean signal strength, T.sub.d.sup.i is the
delay spread, T.sub.dref is a reference time dispersion, K.sub.Td
is a (typically negative) weight factor, e.g., a constant, that
reflects how much the time dispersion affects the cell quality
metric, T.sub.c.sup.i is the coherence time, T.sub.cref is a
reference coherence time, K.sub.Tc is a (typically negative) weight
factor, e.g., a constant, that reflects how much the coherence time
affects the cell quality metric, C.sub.a.sup.i is the antenna
correlation, C.sub.aref is a reference antenna correlation (e.g.,
unity), and K.sub.Ca is a (typically negative) weight factor, e.g.,
a constant, that reflects how much the antenna correlation affects
the cell quality metric.
[0047] It is currently believed that these functions and weights
should be chosen such that the SS.sup.i has the largest impact on
Q.sup.i, and T.sub.d.sup.i, T.sub.c.sup.i, and C.sub.a.sup.i should
bias this value to enable refined selection between cells having
similar SS.sup.i. It can be seen that a time dispersion of
T.sub.dref yields a bias of K.sub.Td, and a time dispersion of zero
yields no bias. A coherence time of T.sub.cref yields a zero bias,
and a coherence time of zero yields a bias of K.sub.Tc. A
correlation of zero yields no bias, and a correlation of C.sub.aref
yields a bias of K.sub.Ca.
[0048] The cell quality metric Q can be expressed by the following
more general relation:
Q.sup.i=g(S.sup.i, T.sub.d.sup.i, F.sub.d.sup.i, C.sub.a.sup.i)
for each cell i, and then the cell selection function f is
just:
Cell=Arg max{Q.sup.i}.
which is described above and which selects that cell having the
largest value of the metric Q. The selector 416 can determine the
largest value Q with, for example, a comparator, and according to
the largest value, the selector 416 may trigger a change of cell
either by itself or by reporting its values to another device in
the network. It will be understood that a cell can be selected in
many alternative but still mathematically equivalent ways to the
cell selection functions f described above.
[0049] Rather than SS measurements, it should be understood that
SIR measurements can be used for cell selection. In such an
apparatus 400, the SS estimator 408 estimates a SIR.sup.i for each
cell, and SIR.sup.i instead of S.sup.i are used by the selector 416
in carrying out the cell selection procedure. In general, the
signal part S.sup.i (the numerator) of the SIR.sup.i can be
estimated as described above, and the interference part I.sup.i
(the denominator), which may reflect either or both of intra-cell
and inter-cell interference, can be estimated using well known
Interference estimation techniques. For example, the estimator 408
can model the received signal per pilot Y.sup.i with the following
equation:
Y.sup.i=h.sup.ip.sup.i+e.sup.i
where h.sup.i is the impulse response of the channel in cell i,
p.sup.i is the pilot symbol, and e.sup.i represents noise. An
estimate h.sub.i of the channel impulse response h.sup.i can be
computed in any of many well known ways, for example as described
in U.S. Patent Application Publication No. 2005/0105647 by
Wilhelmsson et al. for "Channel Estimation by Adaptive
Interpolation". The residual error .sub.i can then be estimated
according to the following expression:
.sup.i=Y.sup.i-h.sup.ip.sup.i
and an estimate of the interference I.sup.i can be determined by
finding the mean over a number of samples | .sup.i|.sup.2.
[0050] The estimates of the signal strength S.sup.i or
signal-to-interference ratio SIR.sup.i, coherence bandwidth
B.sub.c.sup.i, delay spread T.sub.d.sup.i, Doppler spread
F.sub.D.sup.i, antenna correlation C.sub.a.sup.i, etc. can be
generated in an apparatus 400 that is included in a UE as depicted
in FIG. 4A, but this is not necessary. Rather than have the UE
generate a cell selection as shown in FIG. 4A, the UE can send
information to a BS via one or more suitably formatted report
messages, and the BS or other suitable network entity can use the
reported information to generate a cell selection as described
above. The information that the UE would send to the BS would
include at least the signal strength S.sup.i or
signal-to-interference ratio SIR.sup.i.
[0051] For example, the UE can estimate the signal strength S.sup.i
and the antenna correlation C.sub.a.sup.i and send both pieces of
information on a regular basis to a BS or relay node. The BS or
other node would use that information alone or possibly together
with its own estimates of one or both of the frequency and time
selectivity measures (e.g., coherence bandwidth B.sub.c.sup.i,
delay spread T.sub.d.sup.i, Doppler spread F.sub.D.sup.i, etc.) on
the UL channel to generate a cell selection. It is currently
believed that this should pose little difficulty when the duplex
distance, i.e., the frequency difference between the UL and DL, is
low.
[0052] FIG. 4B is a block diagram of such an arrangement, in which
the apparatus 400 is distributed between a UE 401 and another
entity in the communication system, for example, a base station
421. As in FIG. 4A, only some parts of the UE 401 and base station
421 are shown for simplicity. The apparatus 400 again includes the
signal strength estimator 408 and, in this embodiment, the antenna
selectivity estimator 414 in the UE 401 and the cell selector 416
in the base station 421. In FIG. 4B, signals transmitted by the
base station are received by the UEs antenna 402, are
down-converted to base-band signals by a suitable front-end
receiver (Fe RX) 404, and are provided to the estimators 408, 414.
As described above, the estimator 408 generates on a regular basis
for each detected cell i an estimate of at least one of the cells
respective signal strength S.sup.i or signal-to-interference ratio
SIR.sup.i. The estimator 414 is configured to generate estimates of
a measure of the antenna selectivity of the DL channel. Both
estimates are provided to the UEs modulator 418 and FE TX 420 that
up-converts or otherwise transforms the modulation signal for
transmission to the base station. The UE's estimates are received
by a suitable antenna 422 at the base station 421, down-converted
to base-band signals by a suitable Fe RX 424, and recovered by a
detector 426. The recovered UE estimates are provided to the cell
selector 416, and as described above, the cell selector 416
generates a cell selection signal based on the signal strength or
SIR estimate and the antenna selectivity. This cell selection can
then be communicated in a suitable form to the UE 401 through a
modulator 428 and Fe TX 430 in the base station. It will be
understood that other variations are also possible, including for
example a UE in which one or both of frequency and time selectivity
estimators 410, 412 are provided as described above.
[0053] Thus, the apparatus 400, which may be disposed in the UE or
distributed between the UE and another entity of the communication
system, would carry out a method of cell selection such as that
illustrated by the flow chart of FIG. 5. A cell's respective signal
strength S.sup.i or signal-to-interference ratio SIR.sup.i is
estimated (step 502). A measure of at least one of the frequency,
time, and antenna selectivities of the DL channel is estimated
(step 504). For example, one or more of the coherence bandwidth
B.sub.c.sup.i, the delay spread T.sub.d.sup.i, the Doppler
frequency shift F.sub.D.sup.i, and the antenna correlation
C.sub.a.sup.i is estimated as described above. In a distributed
apparatus, these estimates are communicated from a UE to another
system entity, such as a base station (step 506), but as described
above this is not always necessary as indicated by the dashed lines
in FIG. 5. The signal strength is weighted according to the
estimate or estimates of selectivity or selectivities (step 508),
for example by computing a cell quality metric as described above.
These steps are repeated for each cell (step 510), and a cell is
selected (step 512) according to a cell selection function, for
example by determining the largest weighted signal strength as
described above. If needed in a distributed apparatus as indicated
by the dashed lines, the cell selection can be communicated to the
UE (step 514). It will be understood that the order of these steps
and their details can be varied appropriately.
[0054] The cell selection methods and apparatus described above
provide more efficient and robust cell selection that results in
higher QoS, higher capacity, and higher network coverage. To
facilitate understanding, many aspects of this invention are
described in terms of sequences of actions that can be performed
by, for example, elements of a programmable computer system. It
will be recognized that various actions could be performed by
specialized circuits (e.g., discrete logic gates interconnected to
perform a specialized function or application-specific integrated
circuits), by program instructions executed by one or more
processors, or by a combination of both. Wireless receivers
implementing embodiments of this invention can be included in, for
example, mobile telephones, pagers, headsets, laptop computers and
other mobile terminals, and the like.
[0055] Moreover, this invention can additionally be considered to
be embodied entirely within any form of computer-readable storage
medium having stored therein an appropriate set of instructions for
use by or in connection with an instruction-execution system,
apparatus, or device, such as a computer-based system,
processor-containing system, or other system that can fetch
instructions from a medium and execute the instructions. As used
here, a "computer-readable medium" can be any means that can
contain, store, communicate, propagate, or transport the program
for use by or in connection with the instruction-execution system,
apparatus, or device. The computer-readable medium can be, for
example but not limited to, an electronic, magnetic, optical,
electromagnetic, infrared, or semiconductor system, apparatus,
device, or propagation medium. More specific examples (a
non-exhaustive list) of the computer-readable medium include an
electrical connection having one or more wires, a portable computer
diskette, a random-access memory (RAM), a read-only memory (ROM),
an erasable programmable read-only memory (EPROM or Flash memory),
and an optical fiber.
[0056] Thus, the invention may be embodied in many different forms,
not all of which are described above, and all such forms are
contemplated to be within the scope of the invention. For each of
the various aspects of the invention, any such form may be referred
to as "logic configured to" perform a described action, or
alternatively as "logic that" performs a described action.
[0057] It is emphasized that the terms "comprises" and
"comprising", when used in this application, specify the presence
of stated features, integers, steps, or components and do not
preclude the presence or addition of one or more other features,
integers, steps, components, or groups thereof.
[0058] The particular embodiments described above are merely
illustrative and should not be considered restrictive in any way.
The scope of the invention is determined by the following claims,
and all variations and equivalents that fall within the range of
the claims are intended to be embraced therein.
* * * * *